On 09/09/2016 10:26 PM, Alex Rousskov wrote:
On 09/09/2016 11:21 AM, Christos Tsantilas wrote:
On 09/09/2016 07:00 PM, Alex Rousskov wrote:
On 09/09/2016 07:34 AM, Christos Tsantilas wrote:
On 09/09/2016 02:21 PM, Amos Jeffries wrote:
Also the IndependentRunner::registerRunner() method is not used
anywhere. Was it supposed to be called by the child classes ?
(IdleConnList and ConnStateData)
Well, the registerRunner() method should be used instead of
RegisterRunner for ConnStateData and IDleConnList. But this change
somewhere lost while I was playing with the patches.
[rant] Which means our testing was inadequate again. Accidents happen,
but such a major bug should not have went unnoticed if proper testing
was in place.[/rant]
But it is not exactly bug. The registereRunner() method just calls
"RegisterRunner(this)", it does not do anything else.
Just we did not replace RegisterRunner(this); with the
IndependedRunner::registerRunner call.
Ah, this is minor, and no functionality testing could expose this. I
incorrectly thought that we are not registering independent runners at all.
BTW, you can probably move the sketched RegisterRunnerUnlessShutdown()
function into the IndependedRunner::registerRunner() itself and also add
a Must(not an independent runner) check to the sketched
RegisterOrGetRidOfRunner().
I move the code of sketched RegisterRunnerUnlessShutdown to
IndependentRunner::registerRunner()
Also I did not rename the RegisterRunner() to
RegisterOrGetRidOfRunner(), to avoid changes in places other than
runnersRegistry.[cc,h] files..
Now there is Must to assure that the RegisterRunner will not called to
register an IndependentRunner.
The final commit is the patch r14835.
Thank you,
Alex.
_______________________________________________
squid-dev mailing list
squid-dev@lists.squid-cache.org
http://lists.squid-cache.org/listinfo/squid-dev
